A quantitative measure of deviation of a random value distribution from the 'normal' symmetric Gaussian distribution. The skewness is zero for symmetric distributions, it becomes negative (positive) if the distribution contains outliers to the left (right). See higher-order statistics